Annales Geophysicae (2002) 20: 679–690c© European Geophysical Society 2002Annales

Geophysicae

MF radar observations of mean winds and tides over Poker Flat,Alaska (65.1◦ N, 147.5◦ W)

P. Kishore1, S. P. Namboothiri1, K. Igarashi1, Y. Murayama1, and B. J. Watkins2

1Communications Research Laboratory, Tokyo, Japan2Geophysical Institute, Univ. of Alaska, Fairbanks, Alaska, USA

Received: 28 June 2001 – Revised: 1 October 2001 – Accepted: 2 October 2001

Abstract. MF radar wind measurements in the mesosphereand lower thermosphere over Poker Flat, Alaska (65.1◦ N,147.5◦ W) are used to study the features of mean winds andsolar tides. Continuous observation with the newly installedradar is in progress and in the present study we have analyzeda database of the first 27 months (October 1998–December2000) of observation. The observed mean wind climatol-ogy has been compared with previous measurements and thelatest empirical model values (HWM93 model). Similarly,the tidal characteristics are described and compared with theGlobal Scale Wave Model (GSWM00).

The mean wind characteristics observed are fairly con-sistent with previous wind measurements by the Poker FlatMST radar. The main feature of the zonal circulation isthe annual variation with summer westward flow and win-ter eastward flow. The annual mean zonal wind has a west-ward motion at altitudes below 90 km. The annual meanmeridional circulation has mainly southward motion at 70–100 km. There is very good agreement between the radarzonal winds and the HWM93 model winds. Comparison ofthe meridional winds shows some discrepancy. Analysis oftwo years of data indicated that the year-to-year consistencyis preserved in the mean circulation in the mesosphere.

Tidal characteristics observed are also consistent with pre-vious measurements. Semidiurnal tides have the largestamplitudes in summer while the weakest amplitude is ob-served during the winter months. The vertical wavelengthis longer during the summer season compared to the winterseason. Comparison with the GSWM00 produces mixed re-sults. There is reasonable agreement between the observedand modeled phases. Diurnal tide amplitudes are compara-ble in magnitude with that of the semidiurnal tide. Seasonalvariation is less evident in the amplitudes. Comparison of theobserved tidal parameters with the GSWM00 reveals someagreement and discrepancies.

Correspondence to:S. P. Namboothiri(nambooth@crl.go.jp)

Key words. Meteorology and atmospheric dynamics (cli-matology; middle atmosphere dynamics; waves and tides)

1 Introduction

Studies of the arctic middle atmosphere over Poker Flat,Alaska, took a new dimension with the joint research projectrecently launched by the Communications Research Labo-ratory of Japan and the Geophysical Institute of Universityof Alaska, USA. The motivation behind the “Alaska Project”was the complete understanding of the dynamics of the arc-tic middle atmosphere and ionosphere that contribute to theglobal environmental and climate changes. It has been foundthat several of the arctic atmosphere phenomena have directimpacts on the structure or variations of atmosphere overother regions. The main focus of the project is the obser-vation of the middle atmosphere as well as the ionosphereby using various ground-based instruments. As part of theproject various optical and radio equipments have been in-stalled and currently the project is undergoing its observa-tional phase. Mesosphere and lower thermosphere (MLT)observations (in the Alaska Project) are mainly conducted byusing a medium frequency (MF) radar, a Fabry-Perot inter-ferometer, a sodium lidar and a Rayleigh lidar. It is expectedthat the combined use of the data based on these collocatedinstruments will be able to address a number of issues of thedynamics of the high-latitude middle atmosphere.

Information on high-latitude middle atmosphere dynamicsis documented through observations at a number of stationsin the northern and southern hemispheres. At Poker Flat,previous observations using the MST radar during 1980 swere useful to describe various features of the dynamics ofthe middle atmosphere. The radar relied on backscatter dueto turbulent fluctuations in the index of refraction. In ad-dition, meteor echoes have been observed on the radar andhave been used to study the mean winds and tides in the 80–

680 P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska

Fig. 1. Average number of wind esti-mates available per hour at each heightgate (between 50 and 100 km) for sum-mer, winter, spring, and fall seasonsduring 1999. Maximum number possi-ble is 20 estimates/hour.

100 km height range (Avery et al., 1983, 1989; Balsley andRiddle, 1984; Tetenbaum et al., 1986). Various other analy-ses also provided valuable information on other phenomenasuch as gravity waves, turbulence, and longer-period motionsin the high-latitude mesosphere (Carter and Balsley, 1982;Williams and Avery, 1992).

In this paper, we present an analysis of the wind field overPoker Flat for the first 27 months (October 1998–December2000) of MF radar observation. The study is primarily con-cerned with the mean winds and the tidal oscillations in thewind field. The observed results are compared with the lat-est model results as well as with other earlier observations.We begin by describing the data collection and analysis pro-cedures in Sect. 2. Results and discussion are presented inSect. 3. The summary of the analysis is presented in the lastsection.

Table 1. Poker Flat MF radar system specifications

Characteristic

Location 65.1◦N, 147.5◦EOperation frequency 2.43 MHzPeak envelope power 50 kWHalf-power pulse width 27µsecAntenna spacing 170 mSampling interval 2 kmTime resolution 3 min

2 Observations and analysis of data

A new MF radar which uses the spaced-antenna (SA)technique was established at Poker Flat, Alaska (65.1◦ N,

P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska 681

Fig. 2. Time-height cross section ofthe mean zonal wind component (bot-tom) observed at Poker Flat from Octo-ber 1998 to December 2000. HWM93(Hedin et al., 1996) model contours forPoker Flat are given in the top panel.Shaded regions represent the regimes ofwestward flow.

147.5◦ W) in October 1998. Since then, continuous obser-vation of mesosphere lower thermosphere winds has beenunderway. The radar is almost identical to the two otherMF radars installed by CRL, in Yamagawa and Wakkanai inJapan (for details see Namboothiri et al., 2000). Table 1 givesthe specifications of the radar. The radar transmits circularlypolarized radiowaves at a frequency of 2.43 MHz, which isslightly larger than the frequency (∼1.95 MHz) used at Yam-agawa and Wakkanai. Three sets of crossed dipoles, spacedat 170 m, are used for reception. The radar uses a pulselength of∼4 km but data are oversampled at 2 km height in-tervals and observations taken every 3 min. The analysis iscarried out in real-time using the conventional full correlationanalysis (FCA) (Briggs, 1984). Seasonal characteristics arestudied by classifying the data into winter (December, Jan-uary, and February), spring (March, April, and May), sum-mer (June, July, and August), and fall (September, October,and November).

One significant point concerning the data collection overPoker Flat is the availability of echoes even from altitudes

below 50 km. Figure 1 shows the average number of windvalues per hour at each height gate (between 50 and 100 km)for the four seasons of 1999. It should be noted that all ofthe data presented in the study are plotted against the “vir-tual height” of reflection. In a recent study, Namboothiriet al. (1993) described the effects of group retardation on2.2 MHz signals at Saskatoon and predicted the real heightsfor different seasons and solar activity conditions. Theyshowed that noontime group retardation is not severe dur-ing winter for both solar maximum and minimum conditions;for such periods the tidal and winds data are valid up to111 km without any correction. In summer, however, day-time group retardation is more significant and suggest that,during solar minimum periods, the virtual height may beequivalent to the real height up to 97 km. The correspond-ing solar maximum height falls as low as 95 km. Gener-ally, similar constraints are applicable to the Poker Flat data.The results presented here are for the period October 1998–December 2000, which was a solar moderate/maximum pe-riod. From the figure it can be seen that the data yield is

682 P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska

Fig. 3. As in Fig. 2 but for the merid-ional wind. Periods of equatorward mo-tions are denoted by shaded regions.

low at 50–60 km during nighttime. However, the data yieldis much better, even in the altitude range 50–60 km, duringdaytime. There is a clear seasonal variation in data yield atthe lower altitudes where, during the summer, the time ofday is larger than during winter. Generally, the maximumdata rate is observed in the 70–80 km range. The winterseason is characterized by the largest number of wind val-ues; a comparatively lesser data rate is observed in the fallseason. The increased data availability in the altitudes near50 km is due to enhancements in electron density, which maybe the effect of aurora and energetic particle precipitation inthe polar latitudes. Igarashi et al. (2000) reported larger D-region electron density distributions (compared to the IRI-model and Wakkanai, a mid-latitude station) over Poker Flat.The low altitude echoes are also observed in southern hemi-spheric high latitudes. MF radar observations at Scott Base,Antarctica (77.9◦ S, 166.8◦ E) showed some evidence of ion-ization, even at altitudes between 40–50 km, and it is sug-gested that the observed ionization is caused by relativisticelectrons penetrating into the upper stratosphere (von Biel,

1992, 1995). Rees (1963, 1989) discusses the production ofionization from such an electron flux in connection with au-roral ionization.

Harmonic analysis has been performed to derive the pre-vailing wind and tidal components. Hourly means of windsfor each height bin were estimated by averaging all the avail-able wind values for each hour. These hourly mean wind esti-mates for each height gate are averaged to make the monthlyor the seasonal mean of mean winds. For the tidal fitting,we adopt the criterion that there should be at least 16 hoursrepresented in the data for a given height. Also, we weightedthe hourly means in the fit according to the number of valuestherein. Time series of amplitudes and phases were deter-mined and studied the climatological aspects. Consideringbetter data reliability, the analysis is mainly focused on theheight range 60–100 km.

P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska 683

Fig. 4. The observed and modeled wind components for the year 1999. The data are given for 60–100 km (4 km interval). The correlationcoefficient is given in each box.

3 Results and discussion

3.1 Mean winds

This section presents measurements of the mean circulationin the MLT region over Poker Flat and compares them withthe HWM93 empirical model (Hedin et al., 1996). In Fig. 2,the bottom panel illustrates the time-height contours repre-senting average prevailing zonal wind patterns for the first 27months of MF radar observations at Poker Flat. The HWM93empirical model winds for the latitude and longitude of PokerFlat are given in the top panel. Monthly mean values are uti-lized to develop these mean wind climatologies.

Describing the circulation characteristics, the dominantfeature observed is the annual cycle of summer westwardand winter eastward winds. The seasonal variation of the cir-culation is observed mainly at altitudes below 92 km. Twoyears of wind data presented here show qualitative agree-ment in the flow patterns. Generally, the spring reversal fromeastward to westward occurs in March. The summer meso-spheric westward jet maximizes near 75 km at about 35 m/s.This is almost consistent with the past MST radar measure-ments at Poker Flat (Carter and Balsley, 1982; Tetenbaum etal., 1986; Williams and Avery, 1992). The analysis of fiveyears (1985–1989) of wind data by Tetenbaum et al. (1986)

684 P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska

Fig. 5. Annual mean zonal and meridional winds at Poker Flat forthe years 1999 and 2000.

has shown that the summer westward jet has a maximum am-plitude of 30 m/s located at about 80 km. At 80 km the sum-mer westward winds persist for about 6–7 months. The west-ward jet maximizes in June or July. The observed similarity,in the strength of the summer westward jet, with the previ-ous MST radar measurements suggests that the interannualvariability is less evident in the amplitude of the winds. Thepeak summer crossover occurs at around 90 km. Above thisaltitude the wind flow is mostly eastward at all times of theyear. The summer to winter circulation commences duringSeptember-October. Winter eastward winds appear to havehigher variability. The maximum amplitude of the eastwardjet is 20–25 m/s. The eastward maximum tends to occur at alower altitude (around 60 km). This is somewhat different tothe behavior at midlatitude sites, where the maximum east-ward winds occur at an altitude around 75 km. The spring-time seasonal changes show a more abrupt transition in 2000than compared to the transition in 1999. The fall transitionwas comparatively gradual in 1999 and 2000. It occured overa period of several weeks in the 60–90 km height region.

Comparison of the observed zonal winds and the HWM93model winds reveals excellent agreement. The summer andwinter cells have similar structure. The summer westward jethas approximately similar intensities both in model and ob-servation. The maximum difference observed is only 5 m/s.The time occurrence of maximum westward winds is gen-erally the same in both datasets. The peak summer reversaloccurs around 92 km in the model outputs as well as in theobservation. The winter eastward winds also have identicalstrength in both model and observation.

The time-height contours of the meridional wind arepresented in Fig. 3 along with the corresponding model(HWM93) winds. The general pattern of the plots is some-what similar, but there are also some differences. The vari-ability of meridional winds is larger than the zonal winds.

The maximum intensity of southward or northward windsin the model and observation appears to be similar. How-ever, there is some difference in the structure of these south-ward/northward cells. At altitudes above 92 km, the circu-lation shown by the model and observation differ in theirpattern. At 88 km the observation shows a consistent strongsouthward winds (>8 m/s) during June/July. The model pro-duces only weak winds at the same height and time of theyear. These summer equatorward motions have also beenobserved at other high latitude stations, for example, HeissIsland, Molodezhnaya, Tromsø, and Scott Base (Lysenko etal., 1979; Fraser, 1984; Manson et al., 1992). The maximumamplitude of southward winds in the observation and modelis 12–16 m/s and the corresponding northward wind ampli-tude is only 4 m/s. Comparison of the observed meridionalwinds with the earlier MST radar measurements (Tetenbaumet al., 1986) notes some differences. The MST observationsbasically indicated northward motions at altitudes 80–88 kmthroughout the entire year. Our observations show evidenceof southward motions at similar heights during the summerseason. As in the case of zonal winds, the MF and MST radarmeasurements produce meridional winds of comparable in-tensity. The observed features of the summer southward jetare consistent in both years. The location (88 km) of the jetis just below the zonal reversal height where the zonal shearis largest.

Comparing the MF radar mean winds at Poker Flat withthe earlier Mawson (68◦ S, 63◦ E) measurements (Mansonet al., 1991, Figs. 7–8), it can be seen that there is reason-able agreement between the two circulations. At both placesthe zonal winds are of similar strengths and also the circula-tion features have considerable resemblance. Similarly, themeridional contours also compare well in their structure andintensities of the southward/northward cell. The paper byManson et al. (1991) extensively discussed and compared thewind contours at the high-latitude stations such as Poker Flat,Tromsø, Mawson etc. It is shown that the circulation at thesestations is quite similar.

In order to facilitate the comparison more efficiently, wehave calculated the correlation coefficients between the ob-servation and the model values. In Fig. 4 we show the ob-served winds and the model values for altitudes (4 km inter-val) from 60 to 100 km for the year 1999. The correlationvalues are given in each box. It can be seen that the corre-lation is better in the case of zonal winds. The values arearound 0.80 up to 88 km and above this altitude correlationbecomes weak. However, the meridional winds show littlecorrelation with the model values throughout the entire alti-tude. Again, the lower altitudes show comparatively bettercorrelation than the upper altitudes.

The annual mean zonal and meridional profiles for PokerFlat are presented in Fig. 5. It follows from the figure thatthe average zonal wind has a westward circulation at alti-tudes below 90 km and above this limit eastward circula-tion prevails. The westward flow maximizes around 78 kmat approximately 6–9 m/s. The meridional profile showsmainly southward motion in the altitude region 70–100 km

P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska 685

Fig. 6. Semidiurnal seasonal mean for summer, winter, spring, and fall of the Poker Flat radar data (zonal component) and the GSWM00values. Data shown are for the years 1999 and 2000. Horizontal bars represent the standard errors.

Fig. 7. As in Fig. 6 but for the meridional component.

686 P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska

Fig. 8. Height-time cross section of amplitude of the semidiurnaltide of October 1998–December 2000.

and found that the average flow is less than 3 m/s in the sameheight region.

3.2 Semidiurnal tide

This and the following section present observations of thetidal winds over Poker Flat during 1998–2000. The semid-iurnal and diurnal tides are discussed separately and com-pared with the Global Scale Wave Model 2000 (GSWM00)(M. E. Hagan, Private communication, 2001). The advan-tage of GSWM00 over the previous versions is that it canprovide tidal parameters for each month. The vertical struc-tures of the seasonally averaged semidiurnal tide at PokerFlat for 1999 and 2000, and the GSWM00 values are shownin Figs. 6 and 7. The horizontal bars on the plot representthe month-to-month variability about the average. For thepurpose of comparison, the model results were sampled atheights corresponding to those of the radar data, and then theamplitudes and phases of the semidiurnal tide were averaged

Fig. 9. Two-year averaged vertical wavelengths of the semidiurnaltide at Poker Flat. Vertical wavelengths≥140 km are shown bysymbols along 140 km level.

over 3-month blocks to obtain the seasonal means. It seemsthat each component shows almost year-to-year consistencyin the amplitude and phase structures. The amplitudes ofthe meridional component are comparable with those of thezonal component. Weak seasonal variations of amplitude andphase are seen in both radar data and model. Tidal ampli-tudes are largest (10–12 m/s) in summer and fall seasons.The weakest amplitude (5 m/s) occurs in spring. Overall,the comparison between the observation and the model is notgood, especially in the case of amplitudes. Generally, above90 km the model overestimates the observed values. Below90 km, observed values show larger strength than the modelvalues. In almost all the seasons the model amplitude contin-ues to grow with height. Above 80–90 km an abrupt growthin model amplitude is observed. Model amplitudes of values>30 m/s are not shown in the plot. Model profiles show thesmallest amplitude in summer while the other seasons havecomparatively similar values and variations. The phases ofthe zonal and meridional components of the semidiurnal tidegenerally show the downward phase progression during mostof the months. Both model and observation display a largervertical wavelength in summer and a shorter vertical wave-length in winter season. The vertical phase structure gener-ally indicates a propagating wave. Comparison of the modeland observed phases suggests that there is reasonable agree-ment between the two sets of values.

Figure 8 presents contour plots of the zonal and meridionalcomponents of semidiurnal tide amplitudes. The important

P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska 687

Fig. 10. Semidiurnal tidal amplitude and phase observed at PokerFlat for the heights 86 and 96 kms.

feature is the consistent amplitude maximum in the sum-mer/early fall season. The maximum amplitude observed is10–13 m/s. The zonal amplitudes are slightly larger than themeridional component. The winter/spring months are char-acterized by comparatively smaller amplitudes than the am-plitudes in the other months.

The seasonally averaged tidal characteristics reported hereare often consistent with the previous observations made atPoker Flat (Tetenbaum et al., 1986). However, it shouldbe remembered that the comparison with the previous mea-surements is not entirely valid considering the interannualvariability of tides. For example, Tetenbaum et al. (1986)reported semidiurnal tide contours of maximum strength20 m/s for the 1983/1984 data. The present measurementsindicated the value of the maximum contour as 13 m/s. High-latitude tidal structure shows considerable variations be-tween different stations. Tetenbaum et al. (1986) observedlarge differences between the semidiurnal tides at Mawson(MacLeod and Vincent, 1985) and Poker Flat. Avery etal. (1989) compared the high-latitude tidal behavior by using

the data from several stations, including Poker Flat. Theyobserved a number of similarities as well as differences be-tween the stations. We plan to conduct more comparativestudies using the Poker Flat data and concurrent data fromother high-latitude stations.

The two-year (1999 and 2000) average vertical wave-lengths of semidiurnal tides at Poker Flat are shown inFig. 9. Here the height range considered is 80–90 km. Itis clear that long vertical wavelengths (>140 km) or evanes-cent structure are observed during summer and compara-tively shorter wavelengths are observed in other seasons. Thevertical wavelength (zonal) observed in December is verylow (∼13km). It is observed that in both years there are pro-nounced phase jumps which may have contributed to anoma-lous vertical wavelength. Apart from those phase jumps, thegradients in Fig. 6 for winter have reasonably uniform phaseslopes corresponding to vertical wavelengths of about 60 km,comparable with the meridional values. Overall the observedvalues are similar with the MST radar measurements at PokerFlat (Avery et al., 1989).

Figure 10 shows the monthly amplitudes and phases for1999–2000 at two selected heights. This plot illustrates theseasonal variations of tidal parameters at 86 and 96 km. Itcan be seen that the monthly amplitude of the zonal compo-nent correlates well with those of the meridional component.The phases also correlate reasonably well. The zonal phaseleads the meridional phase by about 3 h, indicating that thetide is circularly polarized. As noted in the contour plots theamplitude maximum is observed in late summer/fall months.At both heights, the amplitude is similar in strength. Summerand winter seasons show distinct phase states. The period ofalmost constant phase state (especially at 86 km) in summerlasts longer than the period of constant phase state in winter.Phase transition is observed in spring and fall seasons. Theanalysis of two years of data presented here indicates that theannual pattern of amplitude and phase is repeated in consec-utive years.

3.3 Diurnal tide

Turning our attention to the diurnal tides, Figs. 11 and 12display the mean behavior of height dependencies of tidalparameters for Poker Flat during the four seasons. The dataare presented in the same fashion as the semidiurnal tides(Figs. 6 and 7). First discussing the amplitudes, like thesemidiurnal tide, the diurnal amplitude is also mainly in therange 5–10 m/s. There is no clear seasonal trend evident inthe height profiles. The seasonal profiles shown for the twoyears suggest that the amplitudes appear to have consistentvalues in 1999 and 2000. Both the zonal and meridionalcomponents have roughly similar strengths, especially in thesummer seasons. The zonal amplitude in the winter seasonis marked with slightly larger values than the meridional am-plitude. However, during the spring and fall conditions atsome heights, the meridional component tends to have larger(2–4 m/s) amplitude than the zonal component. A linear dropin the amplitude is evident in both components at 74–94 km

688 P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska

Fig. 11.Height profiles of the amplitudes and phases of the zonal wind component of the diurnal tide during summer, winter, spring, and fallseasons. Data shown are for the years 1999 and 2000. GSWM00 values are also given for comparison.

Fig. 12. As in Fig. 11 but for the meridional component.

P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska 689

Fig. 13. Height-time monthly contours of the amplitudes of thezonal and meridional components of diurnal tide.

during the spring season. The observed amplitudes generallycompare well with the model values during the summer andwinter seasons. Mostly the model values are within the stan-dard deviation of the observed values. However, comparisonis not promising (particularly for the zonal component) inthe case of spring and fall conditions. The model values arelarger than the observed values at altitudes above 75–80 km.

The observed phase profiles show consistency in the twoyears of data analyzed in the present study. The zonal diur-nal tide, during summer and winter, shows downward phasepropagation at altitudes above 80 km. The large error bars forthe meridional component during the winter season demon-strate instability of the phase structure. The spring and fallconditions mainly represent evanescent behavior of the diur-nal tide. It is interesting to note that at many heights (gen-erally below 90 km) the phase of the meridional componentleads that of the zonal component by several hours in all thefour seasons. In spring and fall, sudden changes with height

in the meridional phase are clearly observed. This resultsfrom the interference of different tidal modes. The modelprofiles often show larger vertical wavelengths or evanes-cence in all the seasons. Comparison of the observed andmodel tidal phases does not show much agreement. The dis-crepancy is larger in the case of meridional component.

Figure 13 shows contour plots of amplitude for the zonaland meridional components of diurnal tide at Poker Flat. Thecross sections have some differences compared with the sea-sonal features seen in the previous plots. It appears thatthe zonal and meridional components have the same ampli-tudes. In general, a very distinct seasonal variation is notseen in the climatological presentation of amplitudes. How-ever, above 90 km during summer, there is a tendency to at-tain larger amplitudes for both zonal and meridional compo-nents. The maximum amplitude observed is nearly 10 m/sat around 100 km. Similarly, a closer look reveals that theamplitude is smaller (3–5 m/s) during the winter season. Theyear-to-year variability is least significant in the tidal con-tours.

Comparing the present climatological features with theprevious measurements conducted at Poker Flat (Tetenbaumet al., 1986), some differences can be seen. For example,Tetenbaum et al. (1986) showed that the amplitude of themeridional component is generally larger than the zonal com-ponent. The present MF measurement does not match theprevious measurements. Our measurements agree with themeasurements at Tromsø (Manson et al., 1988), which alsoreported that the zonal and meridional components have thesame amplitudes.

4 Summary

This paper is an attempt to provide the mean wind and tidalcharacteristics over Poker Flat. We have used two years ofdata collected by a new MF radar, which is installed as partof a collaborative research project. Seasonal variations ofmean winds and tides over the site are constructed and theclimatologies are compared with earlier observations as wellas the latest model results.

The main feature of the zonal circulation in the MLT isthe annual variation with summer westward flow and wintereastward flow. The summer westward jet has a maximumintensity of 35 m/s, which is larger than the winter eastwardmaximum (20 m/s). Dominant annual variation, with sum-mer southward motions and winter northward motions, is ob-served in the meridional circulation. The annual mean zonalwind has a westward motion at heights below∼90 km. Itmaximizes at around 80 km with a strength of 7–10 m/s. Theannual mean meridional circulation shows mainly equator-ward motions. The model (HWM93) and observations showvery good agreement (especially for zonal component) in thewind patterns as well as in the strength.

For the semidiurnal tides, comparison between GSWM00results and observations shows that model amplitudes (above80–90 km) are significantly larger than observed amplitudes

690 P. Kishore et al.: MF radar observations of mean winds and tides over Poker Flat, Alaska

during the winter, spring and fall seasons. During the sum-mer, however, observed amplitudes are stronger than themodeled amplitudes. Model and observed tidal phase vari-ations show reasonable similarity. Short vertical wavelengthis observed in winter and evanescent or longer vertical wave-lengths are observed in summer.

The height profiles of diurnal tides suggest weak evidenceof seasonal variation. The summer amplitudes are slightlylarger than the winter amplitudes. The GSWM00 amplitudesand the observed amplitudes match well in the case of diurnaltide. The comparison of the phases shows a larger discrep-ancy particularly in the case of meridional component.

The MF radar observation at Poker Flat is steadily pro-gressing. Creation of an extensive and continuous databaseis a major priority and such a database would open avenuesfor promising collaborative studies with other middle at-mosphere research groups. Topics like studies on latitudi-nal/longitudinal dependencies of major dynamical processesin the MLT region can be well addressed through such initia-tives.

Acknowledgements.We thank M. Hagan for providing the tabu-lations of GSWM00. The authors appreciate the financial sup-port from respective funding agencies. The first author especiallythanks the Telecommunications Advancement Organization (TAO)of Japan for awarding a fellowship. The Poker Flat MF radar isoperated by Communications Research Laboratory in cooperationwith the Geophysical Institute, University of Alaska, Fairbanks,USA.

The Editor in chief thanks D. Thorsen and G. Fraser for theirhelp in evaluating this paper.

References

Avery, S. K., Riddle, A. C., and Balsley, B.: The Poker Flat, Alaska,MST radar as a meteor radar, Radio Sci., 18, 1021–1027, 1983.

Avery, S. K., Vincent, R. A., Phillips, A., Manson, A. H., and Fraser,G. J.: High-latitude tidal behavior in the mesosphere and lowerthermosphere, J. Atmos. Terr. Phys., 51, 595–608, 1989.

Balsley, B. and Riddle, A.: Monthly mean values of the meso-spheric wind field over Poker Flat, Alaska, J. Atmos. Sci., 41,2368–2375, 1984.

Briggs, B. H.: The analysis of spaced sensor records by correlationtechniques, MAP Handbook, 13, 166–184, 1984.

Carter, D. and Balsley, B.: The summer wind field between 80 and93 km observed by the MST radar at Poker Flat, Alaska (65◦ N),J. Atmos. Sci., 39, 2905–2915, 1982.

Fraser, G. J.: Summer circulation in the Antarctic middle atmo-sphere, J. Atmos. Terr. Phys., 46, 143–146, 1984

Hedin, A. E., Fleming, E. L., Manson, A. H., Schmidlin, F. J.,

Avery, S. K., Clark, R. R., Franke, S. J., Fraser, G. J., Tsuda,T., Vial, F., and Vincent, R. A.: Empirical wind model for theupper, middle and lower atmosphere, J. Atmos. Terr. Phys., 58,1421–1447, 1996.

Igarashi, K., Murayama, Y., Nagayama, M., and Kawana, S.: D-region electron density measurements by the MF radar in themiddle and high latitudes, Adv. Space Res., 25, 25–32, 2000.

Lysenko, I. A., Orlyansky, A. D., and Portnyagin, Yu. I.: Results ofwind velocity measurements at middle and high latitudes by themeteor radar method, J. Geomag. Geoelectr., 31, 411–418, 1979.

MacLeod, R. and Vincent, R. A.: Observations of winds in theAntarctic summer mesosphere using the spaced antenna tech-nique, J. Atmos. Terr. Phys., 47, 567–574, 1985.

Manson, A. H., Meek, C. E., Fleming, E., Chandra, S., Vincent, R.A., Phillips, A., Fraser, G. J., Smith, M. J., Fellous, J. L., andMassbeuf, M.: Comparisons between satellite-derived gradientwinds and radar-derived winds from the CIRA-86, J. Atmos. Sci.,48, 411–428, 1991.

Manson, A. H., Meek, C. E., Brekke, A., and Moen, J.: Mesosphereand lower thermosphere (80–120 km) winds and tides from nearTromsø (70◦ N, 10◦ E): comparisons between radars (MF, EIS-CAT, VHF) and rockets, J. Atmos. Terr. Phys., 54, 927–950,1992.

Manson, A. H., Meek, C. E., Hagan, M. E., Hall, C. M., Hocking,W. K., MacDougall, J. W., Franke, S. J., Riggin, D., Fritts, D.C., Vincent, R. A., and Burrage, M. D.: Seasonal variations ofthe semi-diurnal and diurnal tides in the MLT: multi-year MFradar observations from 2 to 70◦ N, and the GSWM tidal model,J. Atmos-Solar. Terr. Phys., 61, 809–828, 1999.

Namboothiri, S. P., Manson, A. H., and Meek, C. E.: E region realheights and their implications for MF radar-derived wind andtidal climatologies, Radio Science, 28, 187–202, 1993.

Namboothiri, S. P., Kishore, P., Igarashi, K., Nakamura, T., andTsuda, T.: MF radar observations of mean winds over Yamagawa(31.2◦ N, 130.6◦ E) and Wakkanai (45.4◦ N, 141.7◦ E), J. Atmos-Solar. Terr. Phys., 62, 1177–1187, 2000.

Rees, M. H.: Auroral ionization and excitation by incident energeticelectrons, Planet. Space Sci., 11, 1209–1216, 1963.

Rees, M. H.: Physics and Chemistry of the Upper Atmosphere (ch.3), Cambridge University Press, Cambridge, 1989.

Tetenbaum, D., Avery, S. K., and Riddle, A. C.: Observations ofmean winds and tides in the upper mesosphere during 1980–1984, using the Poker Flat, Alaska, MST radar as a meteor radar,J. Geophys. Res., 91, 14 539–14 555, 1986.

von Biel, H. A.: Ionisation in the Antarctic stratosphere, J. Atmos.Terr. Phys., 54, 235–242, 1992.

von Biel, H. A.: Ground based radar investigation of the Antarcticmesosphere, Adv. Space Res., 16, 591–598, 1995.

Williams, C. R. and Avery, S. K.: Analysis of long-period waves us-ing the mesosphere-stratosphere-troposphere radar at Poker Flat,Alaska, J. Geophys. Res., 97, 20 855–20 861, 1992.